step of forecasting

19,340 views

Published on

Published in: Technology, Business
0 Comments
4 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
19,340
On SlideShare
0
From Embeds
0
Number of Embeds
5
Actions
Shares
0
Downloads
471
Comments
0
Likes
4
Embeds 0
No embeds

No notes for slide
  • Decisions made by using forecasting technique are more accurate than those made on the basis of “gut” feelings.
  • Most forecast go wrong bcos it is futuristic. Underestimation, oversestimation
  • step of forecasting

    1. 1. Steps Of Forecasting<br />Group 2<br />
    2. 2.
    3. 3.
    4. 4. Determine the use of the forecast<br />Who needs the forecast?<br />All organizations operate in the atmosphere of uncertainty.<br />Decisions to be made affects future of the organization.<br />
    5. 5. Select the items to be forecasted<br />The item to be forecasted.<br /> Dependent variable to be studied.<br />
    6. 6. Determine the time horizon of the forecast<br />Short-range forecast<br />Up to 1 year<br />Purchasing, job scheduling, job assignments<br />Medium-range forecast<br />1 year to 3 years<br />Sales and production planning<br />Long-range forecast<br />3+ years<br />New product planning, research and development<br />
    7. 7. Select Forecasting approach<br />Qualitative Methods<br />Used when situation is vague and little data exist<br />New products<br />New technology<br />Involves intuition, experience<br />
    8. 8. Quantitative Methods<br />Used when situation is ‘stable’ and historical data exist<br />Existing products<br />Current technology<br />Involves mathematical techniques<br />
    9. 9. Data collection<br /> One of the most difficult and time consuming part of forecasting is the collection of valid and reliable data. Forecast can be no more accurate than the data on which it is based<br />Data can be collected from- primary source and secondary source<br />
    10. 10. Four criteria can be applied to the determination of whether the data will be useful-<br />Data should be reliable and accurate<br />Data should be relevant<br />Data should be consistent<br />Data should be timely <br />
    11. 11. Data Reduction<br />Since available data can be either too much or too less, data reduction is necessary.<br />Decide which data is most complete, valid and reliable to increase data accuracy.<br />Some times accurate data may be available but only in certain historic periods.<br />
    12. 12. Exploring Time Series Data Patterns<br />Horizontal pattern- When data observation fluctuate around a constant level or mean<br />Trend pattern- When data observation grow or decline over an extended period of time <br />Cyclic pattern- When data observation exhibits rises and falls that are not of a fixed period<br />Seasonal Pattern- When data observation are influenced by seasonal factors.<br />
    13. 13. Exploring Data Patterns with Auto correlation Analysis<br />Autocorrelation is the correlation between a variable lagged one or more period itself.<br />It is used to detect non randomness of data<br />To identify an appropriate time series model if data is not random<br />
    14. 14.
    15. 15.
    16. 16. Y= 1704/12 = 142<br />r1 = 843/1474 = .572<br />
    17. 17. Select the forecasting model(s)<br />The most prominently used models are:<br />Exponential smoothing method with 1 or 2 variables.<br />Regression Models<br />Once the model has been judicially selected, its parameters are estimated for model fitting purposes.<br />
    18. 18. Make the forecast<br />Forecast is made for a particular period.<br />
    19. 19. Forecast evaluation<br />Comparing Forecast value with actual historical values.<br /><ul><li>……………………… ^</li></ul>Error : et = yt –y t<br />
    20. 20. Thank you<br />

    ×